Although therapy for esophageal cancer has generated a profusion of research programs and clinical trials, difficult issues and important questions remain to be addressed and answered. The scholarly, balanced review of esophageal cancer by Drs. Ilson and Kelsen explores these issues and questions in a state-of-the art summary of treatments for this disease. The authors hold impressive credentials in this field: almost all clinical research in esophageal cancer for the past two decades owes some debt to Dr. Kelsen and his colleagues at Memorial Sloan-Kettering Cancer Center. In this commentary, we will utilize a few of the key points raised by the authors to suggest a slightly different perspective on approaches to esophageal cancer treatment and directions for future research endeavors.
Is There a "Standard Approach" to Therapy?
The authors reference the exhaustive, widely quoted, but unselective worldwide reviews (through the late 1970s) of Earlem and Cunha-Melo to remind readers that 5-year survival statistics for esophageal cancer are a dismal 4% and 6% for surgery and radiation, respectively.[1,2] They also cite a more recent review of selected surgical series, which found 5-year survival to be 12.5% with operative mortality ranging from 5% to 15%. While the authors carefully delineate the studies that have rendered preoperative radiation alone obsolete, we believe that their description of these modalities as "standard approaches" has begun to sound like the recitation of an outdated mantra.
If we are to improve survival and continue research efforts in this field, single-modality surgical therapy cannot and should not be considered a standard approach for potentially curable esophageal cancer. In the current climate of cost-control and cost-benefit analysis, we cannot accept a standard that offers patients a 12.5% chance for cure and carries a 10% risk for immediate mortality. Yet, these are the generally accepted figures for the standard of surgery alone for patients with esophageal cancer.
Should we abandon new esophageal cancer treatment strategies that include surgery? Quite the contrary. However, we believe that if the designation "standard" in front of surgery were replaced with designations such as "unacceptable" or "inadequate," physicians, patients, and health-care policy analysts would rationally and naturally insist that every patient receiving surgery for potentially curative esophageal cancer be placed into a clinical trial with the aim of finding a treatment that improves survival over surgery alone. Indeed, we could imagine a scenario in which reimbursement for treatment took place only if the patient were entered into a clinical trial.
Similarly, we feel just as strongly that we should discontinue the designation of "standard" for radiation plus chemotherapy. To date, the "significantly" better radiation-plus-chemotherapy arm of a well-conducted, prospective randomized trial resulted in a median survival of 12 months. To adopt such therapy as standard is to accept failure for the overwhelming number of patients with localized esophageal cancer. Therefore, it is our contention that all patients treated for cure with a nonoperative program should be entered into clinical trials with the primary objectives of decreasing local cancer and improving the median survival to a minimum of 3 years. Such trials should prospectively track the cost of therapy in each arm and document quality of life by examining both short- and long-term effects of specific therapies.
The argument regarding the designation of "standard" should not be one of mere semantics. In economic terms, regardless of the therapeutic modality used, the bulk of our current expenditures is for noncurative therapy. Those health-care providers who refuse to participate in trials designed to improve survival in esophageal cancer should begin to feel the pressure of driving costs up, through the institution of a health-care policy that predicates reimbursement upon participation in clinical trials. While this may appear to be a drastic change, we believe that such a policy is currently feasible, as NCI-sponsored intergroup trials (discussed by the authors) utilizing operative and nonoperative approaches are available to clinicians and patients in all regions of the United States.
Selection Criteria for Combined-Modality Therapy
Drs. Ilson and Kelsen describe the seminal combined-modality studies that, in fact, form the basis for much community practice in esophageal cancer. The authors treat us to an erudite, well-referenced discussion of the rationale for approaching the curative therapy of esophageal cancer with systemic chemotherapy. However, their introduction to this topic ends with the admonition, "[The] combined-modality approach should be reserved for those patients at highest risk for death from disease recurrence..." The authors advocate the use of endoscopic ultrasonography to differentiate those who have minimal disease from the majority of patients. Can we justify the routine use of this costly, cumbersome procedure to identify a small fraction who can be saved from the ravages of combined chemotherapy and radiation when Herskovic et al reported only 1 death from therapy (1.6%) in 61 patients? Since 87% to 94% of all patients presenting with esophageal cancer expire from their disease, we contend that all unscreened patients presenting with esophageal cancer are at high risk.
Comparison of Operative and Nonoperative Approaches
Although we contend that labeling surgery as a "standard" therapeutic option has served to retard research efforts and will not alter the natural history of esophageal cancer, Drs. Ilson and Kelsen also ask "whether or not esophagectomy is an obligate part of local disease control?..." Apparently, their response (and that of many investigators in this field) is that this question can be answered only "...in the context of a random assignment trial comparing [surgical] treatment of local regional disease to chemoradiotherapy." While it may be fashionable to state that important questions in cancer treatment must be answered by means of a prospective randomized trial, before the direct comparison of an operative to a nonoperative approach is undertaken, we believe that several scientific and ethical issues must be raised:
1) The known immediate mortality from potentially curative surgery is 5 to 10 times greater than the known immediate mortality from potentially curative chemoradiation. Thus, the treatments offer radically different immediate risks.
2) The 2- and 3-year survival rates reported in nonoperative and operative trials are practically identical, leading authorities in the field of esophageal cancer to assert that chemoradiation is an acceptable alternative to surgery. Therefore, as there is only the remotest possibility that a prospectively randomized comparison between a nonoperative approach and an operative approach will demonstrate a statistically significant improvement in survival for either therapy, do we wish to use our most precious resources--our patients--for such a venture?
3) Directing research dollars toward translational programs aimed at defining molecular changes in esophageal cancer that mitigate sensitivity and/or resistance to current treatments will be more fruitful than trying to prove one inadequate therapy is better than another inadequate therapy.
As we do not believe that rearranging schedules or increasing doses of currently active systemic chemotherapy will significantly enhance survival rates, the search for active new agents against esophageal cancer must continue for good performance-status patients with incurable tumors.
Whether racial factors are important in survival of esophageal cancer has been a relevant research issue. A preliminary report by Streeter et al analyzed the Radiation Therapy Oncology Group (RTOG) combined-modality trial and noted that " when all histologies are combined and treated aggressively [in this protocol]...race is not a statistically significant factor in overall survival." Thus, it is very important that aggressive new treatment strategies be made available to all ethnic and socioeconomic groups in the context of clinical trials.
The unprecedented and unexplained increase in the incidence of adenocarcinoma of the esophagus, noted by the authors, is among the most pressing and perplexing problems facing cancer epidemiologists and clinicians. It is the subject of ongoing trials and will be the focus of much future research in esophageal cancer. The fact that the histologic type of esophageal cancer does not currently alter response to treatment is most probably more a reflection of our unsophisticated therapies than a true comment on the molecular biology or molecular epidemiology of esophageal cancer. A more complete understanding of the genetic differences between adenocarcinoma and squamous cell carcinoma of the esophagus will enable us to develop a better feeling for how to individualize future treatments.
As the authors note, "Future strategies in the treatment of esophageal carcinoma will undoubtedly be based upon advances in the understanding of the biochemistry and molecular biology of the disease." Their discussion touches on the numerous intratumoral oncogenic changes identified in esophageal cancers. Some of these oncogenic changes are known to affect the biology of esophageal cancer, while the biologic significance of others has yet to be elucidated, leaving a fertile field for future exploration.